Using (simplified) petrophysical analysis to predict PV10

Morrow/Springer (Oklahoma) Example

Craig J. Peck Canyon Investments, Inc. January 11, 2004

Problem statement

Client was using geologist-derived porosity and Sw to make completion decisions in Morrow/Springer play in the Anadarko Basin

  • No criteria for pay or HPV minimums was in place.
  • All wells “slim-hole” (4 ¾ inch bits)

Several completions did not payout the completion/frac-job (defects)!

Goal was to halve the number of defects using Six Sigma techniques.

What is Six Sigma?

Six Sigma is a disciplined, data-driven approach and methodology for eliminating defects (driving towards six standard deviations between the mean and the nearest specification limit) in any process—from manufacturing to transactional and from product to service.

Six SD’s are 3.4 defects per million, not obtainable in our business; however, the statistical techniques can produce significant savings.

Conventional Well-by-Well Petrophysical Workflow

Content Image
Conventional Well-by-Well Petrophysical Workflow
* Load and depth-shift data. * Perform environmental corrections. GR and Neutron are most significant. * Establish matrix and shale parameters. * Perform Effective Porosity and Dual Water Sw analysis. * CPI (Computer Processed Interpretation) prepared.

Six-Sigma Project petrophysical workflow

  • Acquire wireline data. This was very difficult due to poor historical data collection and filing methods.
    • Slim hole data was mostly from SLB
    • Some Reeves (BPB) and Baker
  • Determine PV10 values (from reservoir engineers) and completion costs.
  • Develop petrophysical model.
  • Correlate petrophysical results to PV10.
  • Create a statistical predictive model to determine PV10.

Petrophysical model

  • TerraStation software macro code used.
  • Rw determined zone-by-zone, from Static SP.
    • Use SP “spreadsheet” to estimate Rw from SP. Spreadsheet freely distributable, created by Andy May (c. 1993).
    • Requires accurate measures for Rmf: Not calculations if possible.
  • No Environmental corrections performed
  • Use Percentiles to determine GRMA (3rd percentile) and GRSH (95th).
  • Paleozoic rocks, so use a linear VSHGR model. Shale porosity values determined from basin-wide parameters.
  • Use DUAL WATER model to calculate SWE and PHIE.
  • Apply bad hole logic
    • RHOB, CN, CAL
    • Manually Edit pulls
  • Apply Pay cutoffs
    • VSH < 40%, PHIE > 6%, SWE < 60%

Simplified CPI Example

Simplified CPI Example

Predictive Model 1

  • PV 10 as a function of:
    • HPV
    • SSP
    • Ratio of Wet sand to RR
  • Correlation coefficient r = 0.85
  • Geologist hand-derived porosity and Sw only explain 45% of the variance in PV10
Predictive Model 1

Predictive Model 2

  • PV 10 as a function of: HPV
  • Correlation coefficient r = 0.75
Predictive Model 2

Results

  • Petrophysical model along with other factors (number of completion zones, mud log shows) generate an even-higher correlation coefficient.
  • Program in-place since January, 2004.
    • Using predictive model would have resulted in zero defects in first 11 months of ‘04.
      The goal of halving the defect percentage was met.
    • Program implementation required implementing log QC techniques with service companies.
  • Even some pre-deployment success.

Client-operated well

  • Client recommended plugging due to likely water
  • Partners took over well
  • Produced water and gas but without sufficient volumes to be commercial
Client-operated well